Nuclear is the cleanest, safest, most reliable energy we already know how to build

Every kilowatt-hour has a footprint that starts before the plant opens and ends long after it closes. Count all of it, and nuclear sits at roughly 12 grams of CO₂ per kWh: comparable to wind, about a quarter of solar, and a tiny fraction of fossil fuels. That's the IPCC's median across global grids — it varies a little with how the electricity used in mining, enrichment, and construction is generated.

The difference is what happens when the sun sets or the wind dies. Renewables drop to zero; nuclear keeps delivering. France, where nuclear supplies roughly two-thirds of electricity, has among the lowest carbon footprints per kilowatt-hour in Europe, without relying on batteries to cover the gaps.

Climate targets need every zero-carbon tool. Wind and solar alone would force staggering amounts of storage and backup; nuclear, which runs whenever you need it, closes the gap they can't.

Per terawatt-hour of electricity, nuclear causes about 0.03 deaths. Lower than wind, comparable to solar, and roughly 800 times safer than coal. That count includes Chernobyl and Fukushima.

Why so safe? Nuclear plants are engineered with defense in depth: multiple redundant barriers, ceramic fuel, steel cladding, a thick concrete containment dome, each designed to contain a failure of the one before. Modern reactors add passive safety: they shut themselves down and cool themselves with gravity, with no operator action and no external power.

The famous accidents prove the point differently than most people assume. Chernobyl used a unique Soviet design with no containment building and ran an ill-conceived test; that combination cannot exist in a licensed Western plant. Fukushima was hit by a tsunami far beyond anything the plant was designed to withstand. The direct death toll from radiation was zero.

Fission is staggeringly dense. A 1,000 MW reactor runs for a year on about 27 tonnes of uranium; a coal station of the same size burns more than 2.5 million tonnes of coal. A single fuel pellet the size of a gummy bear carries as much usable energy as a tonne of coal.

That concentration shows up on the map. For the same 1,000 megawatts of capacity, a nuclear plant uses about a square mile of land, while a wind farm needs roughly 360 and utility solar around 75.

This is land we don't have to pave, fence off, or fight over. Sparing it matters for nature as much as for politics: the least land-hungry clean energy is the one that disturbs the least habitat.

Capacity factor measures how much of a plant's maximum output it delivers over a year. U.S. nuclear reactors sit near 93%, the highest of any source, fossil or clean.

Wind and solar are cheap but intermittent: a grid built on them needs backup for every still day and every cloudy night, plus storage that is still expensive at scale. Nuclear delivers power whenever you need it, which is why it pairs with renewables instead of competing with them.

A nuclear plant is also built to last. Original 40-year licenses are routinely extended to 80, and modern plants are designed for even longer service.

Every reactor's spent fuel from six decades of U.S. commercial nuclear power would fit in a single football field stacked about ten yards deep. It is solid ceramic, sealed in steel, and sits in concrete casks monitored on-site. It has not harmed a single member of the public.

Compare that with coal: a single large coal plant produces around 100,000 tonnes of ash every year, and coal ash released into the environment carries around 100 times more radiation than a nuclear plant making the same electricity, with most of it simply dumped or ponded. Nuclear waste is the one waste stream that is fully contained and paid for.

Geological disposal, burying it deep in stable rock as Finland is now doing at Onkalo, is a solved engineering problem that politics, not physics, has delayed.

Almost all of a reactor's cost is paid before it opens. After that, uranium is a small fraction of the operating budget, maintenance is modest, and the plant runs for 60 to 80 years or more, far longer than any wind turbine or solar panel.

Existing U.S. nuclear plants are routinely among the cheapest electricity on their grids, which is why utilities keep extending their licenses rather than replacing them. The cost problem is not the technology's physics. It's that nearly every recent build has been a one-off design, built one at a time.

Standardize the design, build in series, and the numbers change dramatically: France built 56 reactors in about 15 years, getting faster and cheaper with each one, and South Korea and China now build reactors on time and on budget the same way.

• Cost and schedule overruns are real. Vogtle in Georgia finished years late at roughly double its budget; Hinkley Point C and Olkiluoto tell the same story. The lesson isn't "nuclear can't work." It's that custom, one-off projects are the worst possible way to build them.
• Policy whiplash kills projects. Plants canceled mid-construction after political reversals (like the U.S. in the 1980s) mean every project starts from scratch, and the cost of regulatory stop-start is baked into every estimate.
• Fear is a cost too. Public perception, however disproportionate to the data, lengthens permitting and siting timelines. That's a political problem with an engineering solution: new designs, honest communication, and demonstrated results.
• Waste siting is stuck. The technical path, deep geological disposal, is solved. The political path, choosing a site, is not. Finland proved it can be done; others are still deciding.

Every one of these is a fixable systems problem. The alternatives, staying on fossil fuels or building the required wind, solar, storage, and transmission instead, carry costs that are equally real and far less discussed.